5-Azacytidine Specifically Depletes Regulatory T Cells (Tregs) in Myelodysplastic Syndrome (MDS) Patients

Abstract

Abstract 787 Introduction:The hypomethylating agent 5-azacytidine (5-azaC) leads to improved survival compared to conventional care regimens in patients with intermediate-2 and high-risk MDS and Acute Myeloid Leukaemia (AML) with less than 30% blasts.The precise mode of action of 5-azaC is uncertain, however a combination of cytotoxicity and demethylation is partly responsible for its anti-leukemic activity. In addition, 5-azaC has a profound effect on immune function and inhibits T cell proliferation and activation, blocking cell cycle in the G0 to G1 phase and decreases the production of pro-inflammatory cytokines, suggesting a possible in vivo and in vitro immunomodulating role that may contribute to its anti-leukemic activity. The aim of this study was to investigate the effects of 5-azaC on different subsets of CD4+ T cells, including regulatory T cells (Tregs) and T helpers (Th1, Th2, and Th17). Patients and methods:Seventy intermediate-2/high risk MDS patients and 10 healthy age matched donors (HDs) were studied. CD4+ and CD8+ T cells subsets (percentages and absolute numbers) were investigated by flow cytometry. All patients have received 5-azaC and peripheral blood samples were collected at diagnosis and after 1, 3, 6, 9 and 12 month from initial treatment. On average 3 samples were collected per patient.In vitro study: 5-azaC was added to pre-stimulated PBMCs from 4 HDs and 3 high-risk MDS patients to facilitate the drug incorporation. After 48 hours of initial stimulation, 5-azaC was added every 24 h up to 96 h on two different concentrations (1 μM and 2 μM). For each timepoint (t0, t+24, t+48, t+72, t+96) cells were stained with CD3, CD4, CD25, CD127 and Foxp3 for Tregs and with CD3, CD4, IFN-γ, TNF-α, IL-4, IL-17 for T helpers after an additional 4 hours stimulation with PMA/Ionomycin.Results: In vivo results:Numbers and percentages of Tregs were significantly higher in patients’ peripheral blood prior to treatment compared to HDs (0.7% v 0.08%, p<0.001 and 1.1 × 107/L v 4.6 × 106/L, p=0.01). However, after 12 months of treatment with 5-azaC the number of Tregs decreased to the normal level. Absolute numbers and percentages of Tregs were also higher in non-responder patients compared to responders after treatment (1.2 × 107/L v 7.3 × 106/L, p=0.01).Although the number of Th1 and Th17 cells did not change significantly following treatment with 5-azaC, the Th1/Tregs and Th17/Tregs ratios were significantly decreased in non-responders (p=0.02), whereas these ratios remained stable in responder patients.In vitro results: There were no changes in the number or frequency of Th1, Th2 or Th17 when 5-azaC (1 μM and 2 μM) was added to patients' T cells. However, numbers and frequencies of Tregs dropped significantly compared to HDs' T cells (p=0.034). The ratio of Th1/Tregs and Th17/Tregs were also higher in patients' treated PBMCs after in vitro 5-azaC.There was a significant decrease in the percentages and numbers of Th1 cells (15.4% v 2.7%, p=0.043 and 1.42 × 103 v 6.44 × 104, p=0.021), Th17 cells (1.01% v 0.07%, p = 0.021 and 4.22 × 103 v 7.2 × 102, p=0.021) Th1/Tregs ratio (79.8 v 1.5, p=0.043), and Th17/Tregs ratio (5.2 v 0.1, p=0.021) in HDs’ PBMCs treated with 2 μM compared to untreated cells. 5-azaC also reduced the absolute numbers of CD4+TNF-α+ T cells (1.07 × 105 v 5.37 × 103, p=0.021) and Th2 (9.24 × 103 v 1.25 × 102, p=0.021) in HDs’ T cells. There was no preferential apoptosis in any subsets of T cells confirmed by Annexin V staining. However, it is interesting to note that the telomere length of Tregs treated with 5-azaC was longer than untreated Tregs, suggesting a decrease of their proliferation. Conclusion:Our study suggests that 5-azaC can induce a significant decrease in the number of Tregs in patients (in vivo and in vitro) and HDs, and therefore creates a pro-inflammatory state, despite a small decrease in the number of Th1 and Th17 cells. These changes are more significant in patients who responded to 5-azaC rather than in non-responders. Surprisingly, our in vitro study suggests that 5-azaC leads to a marked reduction in Tregs. As there is not a Tregs' specific apoptosis following 5-azaC treatment, we speculate that the reduction in Tregs' number is mainly due to de-methylation of transcription factors which leads to conversion of Tregs into other T cell subsets (ie Th1 or Th17). Disclosures:Mufti:Celgene: Research Funding.